Microsoft Data Scientist interview questions

1 hour phone screen with general DS/ML questions followed with full day on site loop. 5 1 hour interviews in total. I didn't really know what to expect going in and got a pretty wide range of questions. Topics involved data structures & algos, experimentation, dealing with dirty data, & lots of statistics/machine learning questions. Overall everyone was really friendly sharp & smart. I also had a few curveball questions thrown my way where the interviewer was trying to see if they could catch me off guard.

The interview process consisted of an in-person interview on my college campus scheduled around the career fair, followed by an onsite interview at the headquarters. In terms of difficulty, the first interview was just to check that I could actually code and debug things, while having a subtle behavioral component as well. The onsite cosisted of 4 consecutive one-hour technical rounds. The difficulty of the questions progressed gradually: leetcode easy -> leetcode medium -> complexity theory / algo analysis -> graph algorithms / low-level memory concepts.